Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
MIME-Version: 1.0
In-Reply-To: <1504095687623-0.post@n3.nabble.com>
References: <1504095687623-0.post@n3.nabble.com>
From: =?UTF-8?Q?Daniel_Jeli=C5=84ski?= <djelinski1@gmail.com>
Date: Wed, 30 Aug 2017 19:05:09 +0200
Message-ID: <CAMrH03JWsda9CZ8fUQ2Y=Vc0bS=TqjB2mOTtiXu5uvC0Tcrw9Q@mail.gmail.com>
Subject: Re: Need help with Row Key design
To: user@hbase.apache.org
Content-Type: multipart/alternative; boundary="001a1145a8dec00d4e0557fb87b1"
archived-at: Wed, 30 Aug 2017 17:05:22 -0000

--001a1145a8dec00d4e0557fb87b1
Content-Type: text/plain; charset="UTF-8"

Have you considered Solr or Elasticsearch?

HBase is a rather poor choice for most UI needs. It is great for running
analytics/reports where all rows need to be scanned, and for random access
by key. User interfaces usually require just a few top records to display,
plus they usually need filtering and sorting on different fields. Solr
excels in these workloads, while HBase would need either secondary indexes
(which are not implemented in HBase, but there are some tools that allow
this), or map/reduce jobs, which are unlikely to satisfy your UI users, as
the take too long to run.
Regards,
Daniel


2017-08-30 14:21 GMT+02:00 deepaksharma25 <dsharma.25@gmail.com>:

> Hello,
> I am new to HBase DB and currently evaluating it for one of the requirement
> we have from Customer.
> We are going to write TBs of data in HBase daily and we need to fetch
> specifc data based on filter.
>
> I came to know that it is very important to design the row key in such a
> manner, so that it effectively uses it to fetch the data from the specific
> node instead of scanning thru all the records in the database, based on the
> type of row key we design.
>
> The problem with our requirement is that, we don't have any specific field
> which can be used to define the rowkey. We have around 7-8 fields available
> on the frontend, which can be used to filter the records from HBase.
>
> Can you please suggest, what should be the design of my row key, which will
> help in faster retrieval of the data from TBs of data?
> Attaching here the sample screen I am referring in this
> <http://apache-hbase.679495.n3.nabble.com/file/t497182/SearchScreen1.png>
> .
>
> Thanks,
> Deepak Sharma
>
>
>
> --
> Sent from: http://apache-hbase.679495.n3.nabble.com/HBase-User-
> f4020416.html
>

--001a1145a8dec00d4e0557fb87b1--